Approximately 180,000 patients per year develop brain metastases in the U.S. and the incidence is rising as patients with metastatic cancer survive longer, due to improved systemic therapies, and highly sensitive MR imaging which can detect small metastatic lesions. Stereotactic radiosurgery (SRS) is a common treatment for patients with a limited number of brain metastases. SRS is logistically convenient and is well tolerated in the short term. However, about 10-20% of treated lesions will eventually develop evolving lesions, and these are evenly split between radiation-induced necrosis and recurrent tumors. Necrosis and recurrence occur within the same time interval after SRS, and are difficult to distinguish on MRI or other standard imaging modalities, such as FDG PET. About 5-10% of cases of suspected necrosis and/or recurrence require neuro-surgical resection in order to alleviate symptoms and to establish the pathology associated with the lesion. The objective of this proposal is to improve the management of patients with evolving lesions following SRS for brain metastases. With a more accurate and earlier diagnosis, based on noninvasive imaging of an evolving lesion, we will be able to offer more appropriate and effective therapies in a more timely and efficient manner. This is particularly important since the treatment of necrotic lesions versus recurrent tumor is quite different. Often the toxicity associated with SRS-induced radio necrosis can be managed conservatively, and thus may reduce the need for surgical intervention and resection, thereby reducing patient morbidity/hospitalization and cost. In contrast, tumor recurrence needs to be managed early and more aggressively, in order to control the lesion, prevent growth, and associated symptoms. Our hypothesis is that 18F-Fluorocholine (18F-FCho) PET imaging will be able to distinguish necrosis from tumor recurrence in lesions that are evolving after SRS. This is a novel study. 18F-FCho has not been extensively or exclusively studied for evaluating evolving brain lesions after SRS. Furthermore, this is the first detailed comparison between the level of 18F-FCho radioactivity in tissue samples resected from the lesion and a detailed pathologic assessment the same tissue samples. This correlative imaging, lesion sampling and pathological study will validate 18F-FCho PET as a clinically useful predictor of post-SRS lesion pathology. We will also acquire dynamic contrast enhancement MRI (DCE-MRI) and magnetic resonance spectroscopy (MRS) pre-operatively in all of the patients as part of standard pre-operative evaluation. These advanced imaging datasets will also be assessed for their ability to distinguish necrosis from recurrence and thereby provide a direct reference within this cohort of patients against which the predictive capability of 18F-FCho may be judged.